Loading...
In January 2002, President George W. Bush signed the No Child Left Behind Act (NCLB) with broad bipartisan support. The law's logic was simple: tie federal funding to standardized test scores in reading and math, and schools would improve. By 2006, the stakes were clear — schools that failed to show 'Adequate Yearly Progress' for two consecutive years faced restructuring, staff replacement, or state takeover. The incentives were enormous: billions in Title I funding hung in the balance. Schools responded rationally. In Texas, one of the first states to implement high-stakes testing, the percentage of instructional time spent on tested subjects rose from 60% to 85% between 2001 and 2007. Art, music, social studies, and recess were slashed. A 2007 Center on Education Policy survey found t...
Popular framing: Teachers and administrators gamed the system instead of teaching.
Structural analysis: When billions in funding ride on a single proxy, schools rationally optimize for the proxy; the test becomes the target and the construct it was meant to measure — learning — quietly erodes. Principal-agent distance between legislators and classrooms plus weak construct validity in the chosen metric produce the cobra effect at national scale.
The popular framing treats this as a measurement problem (pick better tests) or an implementation problem (enforce more carefully), missing that any sufficiently high-stakes single metric will be optimized at the expense of the underlying construct it proxies. Without naming Goodhart's Law as the structural mechanism, reform efforts reproduce the same trap — Race to the Top added teacher evaluation metrics and immediately generated the same gaming dynamics in teacher scoring. The gap matters because the lesson of NCLB isn't 'measure better' but 'understand what happens to any measure once it gains power.'
